Joy of Giving: Evidence from a Matching Experiment with Millionaires and the General Population (with Paul Smeets, Ashley Whillans, and Michael Norton)
How much do people benefit from giving to others? Does the size of the gift matter? How does the joy derived from giving vary between persons? We answer these questions in a unique experiment conducted among large samples of the general population (n = 1,232) and of millionaires (n = 807) in the Netherlands. Two thirds of participants entered a lottery with five prizes worth €100 that could be donated to a charity of their choice. For one third of participants in the match treatment, contributions were doubled by the experimenters. After the donation decision, participants reported their mood. We test whether giving improved mood, whether giving more makes people happier than giving less, and whether these mood effects vary with the match. We thank participants at the 5th Science of Philanthropy Initiative Conference in Indianapolis and at the Association for the Study of Generosity in Economics (ASGE) session of the ASSA meeting in Atlanta for comments and suggestions. The preregistration of the research is here. The conference presentation is here.
Altruism, Warm Glow, and Generosity: A National Experiment (with Mark Ottoni-Wilhelm)
To what extent does altruism influence donations to charity? How does empathy induction and a moral appeal to the principle of care affect altruism and generosity? In this research, joint work with Mark Ottoni-Wilhelm (IUPUI), we try to answer this question. The project received funding from the Science of Philanthropy Initiative.
- We presented the design of the envisioned field experiment at the 2nd SPI Conference, November 7-8, 2014.
- A first pilot experiment was conducted on January 23, 2015 to develop a scale to measure the principle of care as a psychological state.
- A second pilot experiment was conducted on April 16 and 17, 2015, to measure the effect of a manipulation of the principle of care through exposure to the mission statement of Oxfam America either in the form of a text, the text with a still image of donors, or a video in which donors voice the text. We extended the fieldwork on July 16, 2015.
- An overview of the findings of the pilot experiments is in the paper presented at the 14th TIBER Symposium on Psychology and Economics and at the 3d SPI Conference, September 10-11, 2015. Slides here.
- A third pilot experiment was conducted in October and November 2016 to test the budgets among a subsample of participants in the Giving in the Netherlands Panel Survey 2016. Hypotheses were preregistered here. Results were presented at the 86th Southern Economic Association Conference, Washington DC, November 19-21, 2016.
- The field experiment remains to be conducted (Fall 2018).
Trust and Giving (with Pamala Wiepking and Matt Bennett)
Does a culture of trust encourage charitable giving? We study whether the level of generalized trust in a country is correlated with the level of charitable giving, over and above the individual level relationship. Building on an interdisciplinary literature we formulate hypotheses explaining if, and if so why and how, a culture of trust promotes charitable giving. We test these hypotheses using multiple data sets, including data from the CAF World Giving Index, the Global Trust Research Consortium, and the Individual International Philanthropy Database. Our results can help understand if and how cultures of trust promote giving. The paper will be presented at the 47th ARNOVA Conference in Austin, Texas.
Mega-Analysis of Generalized Social Trust (with the Global Trust Research Consortium)
With an application to generalized social trust this paper introduces mega-analysis, a method developed to increase the power and generality of conclusions from empirical data-analyses. A mega-analysis uses the largest possible number of observations of a phenomenon to quantify the strength of its correlates. Characteristics of the data (e.g., time, place, collection mode) and measures (e.g., scale properties) are modeled as separate predictors in addition to substantive predictors (i.e., theoretically derived variables). Specific dependent variables as well as relations between a specific dependent variable and a set of predictors can be mega-analyzed. The benefits of mega-analysis emerge from the analysis of generalized trust, a variable included in a large number of surveys and experiments conducted globally. The mega-analysis of trust shows how sample composition, survey design and choices in statistical modeling affect the conclusions on correlates of trust.
Status update (June 13, 2018): We have now identified over 365 surveys that include at least one question on generalized trust. The current master file includes 3,828,342 observations from 203 surveys in 165 countries, from 1953 to 2017. See the Global Trust Research Consortium website for more details.